PRISMS: No Benefit of tPA in Mild Stroke Without Disability

Sue Hughes            

January 31, 2018

LOS ANGELES — Data from the first randomized trial of thrombolysis in patients with mild ischemic stroke and no clear disabling deficit show no evidence of benefit of tissue plasminogen activator (tPA) and an increased risk for symptomatic intracerebral hemorrhage in this population.

"Although the results of the PRISMS trial are not reliable as the study was stopped early due to slow recruitment and so was underpowered, based on the data we do have it appears unlikely that the benefits of tPA extend to those stroke patients without clearly disabling deficits at presentation," lead investigator, Pooja Khatri, MD, University of Cincinnati, Ohio, told Medscape Medical News.

"This is a big disappointment," she added. "We had hoped that tPA would be beneficial in patients with just a very mild neurological deficit, but we are now a step closer to making an informed decision about this."

She estimated that the PRISMS trial population reflected about one third of ischemic stroke patients. "So we are talking about a large group, and it is important to have evidence-based information on tPA in this group."

Dr Khatri stressed that it was important to understand that this trial excluded patients with a clear disability deficit. "This trial included patients who were still deemed able to carry out basic activities of daily living, such as walking, bathing/dressing, toileting, and eating, and be able to return to work if appropriate. They had some neurological deficit, but this could have been very mild, such as a little facial droop or subtle weakness in the arm. The investigator with the patient’s input had to make a clinical judgment on this based on presenting symptoms."

The results do not apply to patients with a mild stroke who are judged to have a disabling deficit that would affect their basic activities of daily living, she noted. "These patients should receive thrombolysis as this has been shown beneficial in previous studies."

The trial was presented here at the International Stroke Conference (ISC) 2018.

Chair of the ISC 2018 session at which the PRISM trial was presented, Bo Norrving, MD, Lund University, Sweden, said, "By stopping early, the study was underpowered for any firm conclusions, and the confidence intervals were very large, so it’s difficult to draw definitive conclusions. Earlier studies have showed benefit if the patient has mild stroke with symptoms that are disabling. This study included patients with mild symptoms that were not disabling — so, a different cohort — and the results do not support use of tPA in such patients." 

He added, however, that these results were preliminary. "It will be important to have more details about what symptoms these patients had," Dr Norrving said. "An issue is also the reliability of determining in the very acute case if symptoms are disabling or not. It would be good to have more data on the precision of this."

In her presentation, Dr Khatri explained that more than half of all ischemic stroke patients present with mild stroke, defined as National Institutes of Health Stroke Scale (NIHSS) scores of 0 to 5.

Evidence from previous studies supports treating patients with mild stroke who have a clear disability deficit, but exactly which patients with mild strokes to treat "is somewhat of a gray area as different trials have excluded different types of mild stroke," she said.

In the subset of patients with deficits judged as nondisabling, the benefit of thrombolysis is unclear. Few such patients were enrolled in major randomized trials, and guidelines reflect community equipoise for this subset, with a class IIb recommendation (level of evidence C).  

In the PRISMS trial, investigators focused on these patients with NIHSS scores of 0 to 5 and no clear disability deficit. "Previous data have suggested that patients who are thought not to have a disability deficit when they first present can often turn out to have such a disabling deficit later on. So it was thought there could be some benefit to treatment in this group."

In the trial, such patients presenting within 3 hours of last known well were randomly assigned to standard dose of tPA (alteplase, Genentech) or oral aspirin (325 mg) as the control.

The trial was stopped early by the sponsor, Genentech, because the prespecified recruitment targets were not being met. 

"We had enrolled 313 of 948 planned patients by 32 months," Dr Khatri commented. "We don’t know the reasons for the slow recruitment. It is possible that doctors were reluctant to enroll patients as there is a trend to treat these patients, even though there is no good evidence for this."

The intention-to-treat analysis included 154 patients in the tPA group and 153 in the control group.

The primary endpoint — a good functional outcome of 0 to 1 on the modified Rankin Scale (mRS) at 90 days — was seen in 78.2% of the alteplase group vs 81.5% of the controls, giving an adjusted risk difference of –1.10% (95% confidence interval, –9.44% to 7.25%).

Dr Khatri said the 90-day mRS 0 to 1 outcome was better than expected in the control group, with prior literature suggesting this would be nearer 70% than the 81% seen in this trial. 

The primary safety endpoint, symptomatic intracerebral hemorrhage (ICH) within 36 hours, occurred in 5 patients in the tPA group (3.2%) vs 0 in the control group, although there was no associated increase in mortality. Any ICH occurred in 11 tPA patients (7.1%) vs 5 control patients (3.3%).  

"So we appear to be seeing a harm without any suggestion of benefit," Dr Khatri noted. 

 Almost all other secondary or exploratory outcomes were also unfavorable.

For clinical interpretation and future trial planning purposes, the researchers conducted a Bayesian analysis in which the PRISMS unadjusted outcome proportions were added to an uninformative prior. This suggested the likelihood of tPA having any benefit in this population was 23%, and the chances of an absolute benefit of more than 6% (in achieving mRS scores of 0 to 1) was just 1.9%.

Dr Khatri noted the trial had several limitations, the major one being early termination and resulting low power. There was also a relatively high rate of missing 90-day outcomes, but she added that 30-day mRS score is known to permit robust imputation.

She said that there is still a possibility that tPA may be of benefit in higher-risk subsets of the mild stroke population, such as those with a large-vessel occlusion. This is being studied in another ongoing trial known as TEMPO-2.

The PRISMS trial was sponsored by Genentech. The University of Cincinnati Department of Neurology received funds from Genentech towards conduct of the study.

 International Stroke Conference (ISC) 2018. Abstract LB 9. Presented January 25, 2018. 

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.